The mobile in-App service analysis, aiming at classifying mobile internet traffic into different types of service usages, has become a challenging and emergent task for mobile service providers due to the increasing adoption of secure protocols for in-App services. While some efforts have been made for the classification of mobile internet traffic, existing methods rely on complex feature construction and large storage cache, which lead to low processing speed, and thus not practical for online real-time scenarios. To this end, we develop an iterative analyzer for classifying encrypted mobile traffic in a real-time way. Specifically, we first select an optimal set of most discriminative features from raw features extracted from traffic packet sequences by a novel Maximizing Inner activity similarity and Minimizing Different activity similarity (MIMD) measurement. To develop the online analyzer, we first represent a traffic flow with a series of time windows, which are described by the optimal feature vector and are updated iteratively at the packet level. Instead of extracting feature elements from a series of raw traffic packets, our feature elements are updated when a new traffic packet is observed and the storage of raw traffic packets is not required. The time windows generated from the same service usage activity are grouped by our proposed method, namely, recursive time continuity constrained KMeans clustering (rCKC). The feature vectors of cluster centers are then fed into a random forest classifier to identify corresponding service usages. Finally, we provide extensive experiments on real-world Internet traffic data from Wechat, Whatsapp, and Facebook to demonstrate the effectiveness and efficiency of our approach. The results show that the proposed analyzer provides high accuracy in real-world scenarios, and has low storage cache requirement as well as fast processing speed.
{"title":"Effective and Real-time In-App Activity Analysis in Encrypted Internet Traffic Streams","authors":"Junming Liu, Yanjie Fu, Jingci Ming, Y. Ren, Leilei Sun, Hui Xiong","doi":"10.1145/3097983.3098049","DOIUrl":"https://doi.org/10.1145/3097983.3098049","url":null,"abstract":"The mobile in-App service analysis, aiming at classifying mobile internet traffic into different types of service usages, has become a challenging and emergent task for mobile service providers due to the increasing adoption of secure protocols for in-App services. While some efforts have been made for the classification of mobile internet traffic, existing methods rely on complex feature construction and large storage cache, which lead to low processing speed, and thus not practical for online real-time scenarios. To this end, we develop an iterative analyzer for classifying encrypted mobile traffic in a real-time way. Specifically, we first select an optimal set of most discriminative features from raw features extracted from traffic packet sequences by a novel Maximizing Inner activity similarity and Minimizing Different activity similarity (MIMD) measurement. To develop the online analyzer, we first represent a traffic flow with a series of time windows, which are described by the optimal feature vector and are updated iteratively at the packet level. Instead of extracting feature elements from a series of raw traffic packets, our feature elements are updated when a new traffic packet is observed and the storage of raw traffic packets is not required. The time windows generated from the same service usage activity are grouped by our proposed method, namely, recursive time continuity constrained KMeans clustering (rCKC). The feature vectors of cluster centers are then fed into a random forest classifier to identify corresponding service usages. Finally, we provide extensive experiments on real-world Internet traffic data from Wechat, Whatsapp, and Facebook to demonstrate the effectiveness and efficiency of our approach. The results show that the proposed analyzer provides high accuracy in real-world scenarios, and has low storage cache requirement as well as fast processing speed.","PeriodicalId":314049,"journal":{"name":"Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116402638","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We study the problem of representation learning in heterogeneous networks. Its unique challenges come from the existence of multiple types of nodes and links, which limit the feasibility of the conventional network embedding techniques. We develop two scalable representation learning models, namely metapath2vec and metapath2vec++. The metapath2vec model formalizes meta-path-based random walks to construct the heterogeneous neighborhood of a node and then leverages a heterogeneous skip-gram model to perform node embeddings. The metapath2vec++ model further enables the simultaneous modeling of structural and semantic correlations in heterogeneous networks. Extensive experiments show that metapath2vec and metapath2vec++ are able to not only outperform state-of-the-art embedding models in various heterogeneous network mining tasks, such as node classification, clustering, and similarity search, but also discern the structural and semantic correlations between diverse network objects.
{"title":"metapath2vec: Scalable Representation Learning for Heterogeneous Networks","authors":"Yuxiao Dong, N. Chawla, A. Swami","doi":"10.1145/3097983.3098036","DOIUrl":"https://doi.org/10.1145/3097983.3098036","url":null,"abstract":"We study the problem of representation learning in heterogeneous networks. Its unique challenges come from the existence of multiple types of nodes and links, which limit the feasibility of the conventional network embedding techniques. We develop two scalable representation learning models, namely metapath2vec and metapath2vec++. The metapath2vec model formalizes meta-path-based random walks to construct the heterogeneous neighborhood of a node and then leverages a heterogeneous skip-gram model to perform node embeddings. The metapath2vec++ model further enables the simultaneous modeling of structural and semantic correlations in heterogeneous networks. Extensive experiments show that metapath2vec and metapath2vec++ are able to not only outperform state-of-the-art embedding models in various heterogeneous network mining tasks, such as node classification, clustering, and similarity search, but also discern the structural and semantic correlations between diverse network objects.","PeriodicalId":314049,"journal":{"name":"Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113953901","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Existing research on finding social groups mostly focuses on dense subgraphs in social networks. However, finding socially tenuous groups also has many important applications. In this paper, we introduce the notion of k-triangles to measure the tenuity of a group. We then formulate a new research problem, Minimum k-Triangle Disconnected Group (MkTG), to find a socially tenuous group from online social networks. We prove that MkTG is NP-Hard and inapproximable within any ratio in arbitrary graphs but polynomial-time tractable in threshold graphs. Two algorithms, namely TERA and TERA-ADV, are designed to exploit graph-theoretical approaches for solving MkTG on general graphs effectively and efficiently. Experimental results on seven real datasets manifest that the proposed algorithms outperform existing approaches in both efficiency and solution quality.
{"title":"On Finding Socially Tenuous Groups for Online Social Networks","authors":"Chih-Ya Shen, Liang-Hao Huang, De-Nian Yang, Hong-Han Shuai, Wang-Chien Lee, Ming-Syan Chen","doi":"10.1145/3097983.3097995","DOIUrl":"https://doi.org/10.1145/3097983.3097995","url":null,"abstract":"Existing research on finding social groups mostly focuses on dense subgraphs in social networks. However, finding socially tenuous groups also has many important applications. In this paper, we introduce the notion of k-triangles to measure the tenuity of a group. We then formulate a new research problem, Minimum k-Triangle Disconnected Group (MkTG), to find a socially tenuous group from online social networks. We prove that MkTG is NP-Hard and inapproximable within any ratio in arbitrary graphs but polynomial-time tractable in threshold graphs. Two algorithms, namely TERA and TERA-ADV, are designed to exploit graph-theoretical approaches for solving MkTG on general graphs effectively and efficiently. Experimental results on seven real datasets manifest that the proposed algorithms outperform existing approaches in both efficiency and solution quality.","PeriodicalId":314049,"journal":{"name":"Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114933789","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Adopting algorithmic decision-making in a large and complex enterprise such as a Fortune 50 retailer like Target takes much more than clean, reliable data and great data mining capabilities. Yet data practitioners too often start with advanced math and fancy algorithms, rather than working hand-in-hand with business partners to identify and understand the biggest business problems. (Then teams should move onto how algorithms can be applied to those problems.) Another key step for data scientists at large organizations: ensuring that their business partners -- the merchants, marketers and supply chain experts -- have a base-line understanding of advanced models as well as the proper analytical support tools. Obtaining widespread buy-in and enthusiasm also requires providing a user-friendly interface for business partners with optionality and flexibility that allows the intelligence to be applied to the many varied issues facing a modern retailer, from personalization to supply chain transformation to decisions on assortment and pricing. This talk will explore effective practices and processes -- the do's and don'ts -- for data scientists to succeed in large, complex organizations like a retailer with 1,800+ stores, major marketing campaigns across multiple channels and a fast growing online business.
{"title":"It Takes More than Math and Engineering to Hit the Bullseye with Data","authors":"P. Desai","doi":"10.1145/3097983.3105816","DOIUrl":"https://doi.org/10.1145/3097983.3105816","url":null,"abstract":"Adopting algorithmic decision-making in a large and complex enterprise such as a Fortune 50 retailer like Target takes much more than clean, reliable data and great data mining capabilities. Yet data practitioners too often start with advanced math and fancy algorithms, rather than working hand-in-hand with business partners to identify and understand the biggest business problems. (Then teams should move onto how algorithms can be applied to those problems.) Another key step for data scientists at large organizations: ensuring that their business partners -- the merchants, marketers and supply chain experts -- have a base-line understanding of advanced models as well as the proper analytical support tools. Obtaining widespread buy-in and enthusiasm also requires providing a user-friendly interface for business partners with optionality and flexibility that allows the intelligence to be applied to the many varied issues facing a modern retailer, from personalization to supply chain transformation to decisions on assortment and pricing. This talk will explore effective practices and processes -- the do's and don'ts -- for data scientists to succeed in large, complex organizations like a retailer with 1,800+ stores, major marketing campaigns across multiple channels and a fast growing online business.","PeriodicalId":314049,"journal":{"name":"Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123425061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Tingjin Luo, Weizhong Zhang, Shang Qiu, Yang Yang, Dong-yun Yi, Guangtao Wang, Jieping Ye, Jie Wang
Functional annotation of human genes is fundamentally important for understanding the molecular basis of various genetic diseases. A major challenge in determining the functions of human genes lies in the functional diversity of proteins, that is, a gene can perform different functions as it may consist of multiple protein coding isoforms (PCIs). Therefore, differentiating functions of PCIs can significantly deepen our understanding of the functions of genes. However, due to the lack of isoform-level gold-standards (ground-truth annotation), many existing functional annotation approaches are developed at gene-level. In this paper, we propose a novel approach to differentiate the functions of PCIs by integrating sparse simplex projection---that is, a nonconvex sparsity-inducing regularizer---with the framework of multi-instance learning (MIL). Specifically, we label the genes that are annotated to the function under consideration as positive bags and the genes without the function as negative bags. Then, by sparse projections onto simplex, we learn a mapping that embeds the original bag space to a discriminative feature space. Our framework is flexible to incorporate various smooth and non-smooth loss functions such as logistic loss and hinge loss. To solve the resulting highly nontrivial non-convex and non-smooth optimization problem, we further develop an efficient block coordinate descent algorithm. Extensive experiments on human genome data demonstrate that the proposed approaches significantly outperform the state-of-the-art methods in terms of functional annotation accuracy of human PCIs and efficiency.
{"title":"Functional Annotation of Human Protein Coding Isoforms via Non-convex Multi-Instance Learning","authors":"Tingjin Luo, Weizhong Zhang, Shang Qiu, Yang Yang, Dong-yun Yi, Guangtao Wang, Jieping Ye, Jie Wang","doi":"10.1145/3097983.3097984","DOIUrl":"https://doi.org/10.1145/3097983.3097984","url":null,"abstract":"Functional annotation of human genes is fundamentally important for understanding the molecular basis of various genetic diseases. A major challenge in determining the functions of human genes lies in the functional diversity of proteins, that is, a gene can perform different functions as it may consist of multiple protein coding isoforms (PCIs). Therefore, differentiating functions of PCIs can significantly deepen our understanding of the functions of genes. However, due to the lack of isoform-level gold-standards (ground-truth annotation), many existing functional annotation approaches are developed at gene-level. In this paper, we propose a novel approach to differentiate the functions of PCIs by integrating sparse simplex projection---that is, a nonconvex sparsity-inducing regularizer---with the framework of multi-instance learning (MIL). Specifically, we label the genes that are annotated to the function under consideration as positive bags and the genes without the function as negative bags. Then, by sparse projections onto simplex, we learn a mapping that embeds the original bag space to a discriminative feature space. Our framework is flexible to incorporate various smooth and non-smooth loss functions such as logistic loss and hinge loss. To solve the resulting highly nontrivial non-convex and non-smooth optimization problem, we further develop an efficient block coordinate descent algorithm. Extensive experiments on human genome data demonstrate that the proposed approaches significantly outperform the state-of-the-art methods in terms of functional annotation accuracy of human PCIs and efficiency.","PeriodicalId":314049,"journal":{"name":"Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125563651","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Training thin deep networks following the student-teacher learning paradigm has received intensive attention because of its excellent performance. However, to the best of our knowledge, most existing work mainly considers one single teacher network. In practice, a student may access multiple teachers, and multiple teacher networks together provide comprehensive guidance that is beneficial for training the student network. In this paper, we present a method to train a thin deep network by incorporating multiple teacher networks not only in output layer by averaging the softened outputs (dark knowledge) from different networks, but also in the intermediate layers by imposing a constraint about the dissimilarity among examples. We suggest that the relative dissimilarity between intermediate representations of different examples serves as a more flexible and appropriate guidance from teacher networks. Then triplets are utilized to encourage the consistence of these relative dissimilarity relationships between the student network and teacher networks. Moreover, we leverage a voting strategy to unify multiple relative dissimilarity information provided by multiple teacher networks, which realizes their incorporation in the intermediate layers. Extensive experimental results demonstrated that our method is capable of generating a well-performed student network, with the classification accuracy comparable or even superior to all teacher networks, yet having much fewer parameters and being much faster in running.
{"title":"Learning from Multiple Teacher Networks","authors":"Shan You, Chang Xu, Chao Xu, D. Tao","doi":"10.1145/3097983.3098135","DOIUrl":"https://doi.org/10.1145/3097983.3098135","url":null,"abstract":"Training thin deep networks following the student-teacher learning paradigm has received intensive attention because of its excellent performance. However, to the best of our knowledge, most existing work mainly considers one single teacher network. In practice, a student may access multiple teachers, and multiple teacher networks together provide comprehensive guidance that is beneficial for training the student network. In this paper, we present a method to train a thin deep network by incorporating multiple teacher networks not only in output layer by averaging the softened outputs (dark knowledge) from different networks, but also in the intermediate layers by imposing a constraint about the dissimilarity among examples. We suggest that the relative dissimilarity between intermediate representations of different examples serves as a more flexible and appropriate guidance from teacher networks. Then triplets are utilized to encourage the consistence of these relative dissimilarity relationships between the student network and teacher networks. Moreover, we leverage a voting strategy to unify multiple relative dissimilarity information provided by multiple teacher networks, which realizes their incorporation in the intermediate layers. Extensive experimental results demonstrated that our method is capable of generating a well-performed student network, with the classification accuracy comparable or even superior to all teacher networks, yet having much fewer parameters and being much faster in running.","PeriodicalId":314049,"journal":{"name":"Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116722074","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Inci M. Baytas, Cao Xiao, Xi Sheryl Zhang, Fei Wang, Anil K. Jain, Jiayu Zhou
In the study of various diseases, heterogeneity among patients usually leads to different progression patterns and may require different types of therapeutic intervention. Therefore, it is important to study patient subtyping, which is grouping of patients into disease characterizing subtypes. Subtyping from complex patient data is challenging because of the information heterogeneity and temporal dynamics. Long-Short Term Memory (LSTM) has been successfully used in many domains for processing sequential data, and recently applied for analyzing longitudinal patient records. The LSTM units are designed to handle data with constant elapsed times between consecutive elements of a sequence. Given that time lapse between successive elements in patient records can vary from days to months, the design of traditional LSTM may lead to suboptimal performance. In this paper, we propose a novel LSTM unit called Time-Aware LSTM (T-LSTM) to handle irregular time intervals in longitudinal patient records. We learn a subspace decomposition of the cell memory which enables time decay to discount the memory content according to the elapsed time. We propose a patient subtyping model that leverages the proposed T-LSTM in an auto-encoder to learn a powerful single representation for sequential records of patients, which are then used to cluster patients into clinical subtypes. Experiments on synthetic and real world datasets show that the proposed T-LSTM architecture captures the underlying structures in the sequences with time irregularities.
{"title":"Patient Subtyping via Time-Aware LSTM Networks","authors":"Inci M. Baytas, Cao Xiao, Xi Sheryl Zhang, Fei Wang, Anil K. Jain, Jiayu Zhou","doi":"10.1145/3097983.3097997","DOIUrl":"https://doi.org/10.1145/3097983.3097997","url":null,"abstract":"In the study of various diseases, heterogeneity among patients usually leads to different progression patterns and may require different types of therapeutic intervention. Therefore, it is important to study patient subtyping, which is grouping of patients into disease characterizing subtypes. Subtyping from complex patient data is challenging because of the information heterogeneity and temporal dynamics. Long-Short Term Memory (LSTM) has been successfully used in many domains for processing sequential data, and recently applied for analyzing longitudinal patient records. The LSTM units are designed to handle data with constant elapsed times between consecutive elements of a sequence. Given that time lapse between successive elements in patient records can vary from days to months, the design of traditional LSTM may lead to suboptimal performance. In this paper, we propose a novel LSTM unit called Time-Aware LSTM (T-LSTM) to handle irregular time intervals in longitudinal patient records. We learn a subspace decomposition of the cell memory which enables time decay to discount the memory content according to the elapsed time. We propose a patient subtyping model that leverages the proposed T-LSTM in an auto-encoder to learn a powerful single representation for sequential records of patients, which are then used to cluster patients into clinical subtypes. Experiments on synthetic and real world datasets show that the proposed T-LSTM architecture captures the underlying structures in the sequences with time irregularities.","PeriodicalId":314049,"journal":{"name":"Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114050270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Huan Zhao, Quanming Yao, Jianda Li, Yangqiu Song, Lee
Heterogeneous Information Network (HIN) is a natural and general representation of data in modern large commercial recommender systems which involve heterogeneous types of data. HIN based recommenders face two problems: how to represent the high-level semantics of recommendations and how to fuse the heterogeneous information to make recommendations. In this paper, we solve the two problems by first introducing the concept of meta-graph to HIN-based recommendation, and then solving the information fusion problem with a "matrix factorization (MF) + factorization machine (FM)" approach. For the similarities generated by each meta-graph, we perform standard MF to generate latent features for both users and items. With different meta-graph based features, we propose to use FM with Group lasso (FMG) to automatically learn from the observed ratings to effectively select useful meta-graph based features. Experimental results on two real-world datasets, Amazon and Yelp, show the effectiveness of our approach compared to state-of-the-art FM and other HIN-based recommendation algorithms.
{"title":"Meta-Graph Based Recommendation Fusion over Heterogeneous Information Networks","authors":"Huan Zhao, Quanming Yao, Jianda Li, Yangqiu Song, Lee","doi":"10.1145/3097983.3098063","DOIUrl":"https://doi.org/10.1145/3097983.3098063","url":null,"abstract":"Heterogeneous Information Network (HIN) is a natural and general representation of data in modern large commercial recommender systems which involve heterogeneous types of data. HIN based recommenders face two problems: how to represent the high-level semantics of recommendations and how to fuse the heterogeneous information to make recommendations. In this paper, we solve the two problems by first introducing the concept of meta-graph to HIN-based recommendation, and then solving the information fusion problem with a \"matrix factorization (MF) + factorization machine (FM)\" approach. For the similarities generated by each meta-graph, we perform standard MF to generate latent features for both users and items. With different meta-graph based features, we propose to use FM with Group lasso (FMG) to automatically learn from the observed ratings to effectively select useful meta-graph based features. Experimental results on two real-world datasets, Amazon and Yelp, show the effectiveness of our approach compared to state-of-the-art FM and other HIN-based recommendation algorithms.","PeriodicalId":314049,"journal":{"name":"Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121843772","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this talk, I'd like to discuss the intertwining importance and connections of three principles of data science in the title in data-driven decisions. Making prediction as its central task and embracing computation as its core, machine learning has enabled wide-ranging data-driven successes. Prediction is a useful way to check with reality. Good prediction implicitly assumes stability between past and future. Stability (relative to data and model perturbations) is also a minimum requirement for interpretability and reproducibility of data driven results (cf. Yu, 2013). It is closely related to uncertainty assessment. Obviously, both prediction and stability principles can not be employed without feasible computational algorithms, hence the importance of computability. The three principles will be demonstrated in the context of two neuroscience projects and through analytical connections. In particular, the first project adds stability to predictive modeling used for reconstruction of movies from fMRI brain signlas for interpretable models. The second project use predictive transfer learning that combines AlexNet, GoogleNet and VGG with single V4 neuron data for state-of-the-art prediction performance. Our results lend support, to a certain extent, to the resemblance of these CNNs to brain and at the same time provide stable pattern interpretations of neurons in the difficult primate visual cortex V4.
{"title":"Three Principles of Data Science: Predictability, Stability and Computability","authors":"Bin Yu","doi":"10.1145/3097983.3105808","DOIUrl":"https://doi.org/10.1145/3097983.3105808","url":null,"abstract":"In this talk, I'd like to discuss the intertwining importance and connections of three principles of data science in the title in data-driven decisions. Making prediction as its central task and embracing computation as its core, machine learning has enabled wide-ranging data-driven successes. Prediction is a useful way to check with reality. Good prediction implicitly assumes stability between past and future. Stability (relative to data and model perturbations) is also a minimum requirement for interpretability and reproducibility of data driven results (cf. Yu, 2013). It is closely related to uncertainty assessment. Obviously, both prediction and stability principles can not be employed without feasible computational algorithms, hence the importance of computability. The three principles will be demonstrated in the context of two neuroscience projects and through analytical connections. In particular, the first project adds stability to predictive modeling used for reconstruction of movies from fMRI brain signlas for interpretable models. The second project use predictive transfer learning that combines AlexNet, GoogleNet and VGG with single V4 neuron data for state-of-the-art prediction performance. Our results lend support, to a certain extent, to the resemblance of these CNNs to brain and at the same time provide stable pattern interpretations of neurons in the difficult primate visual cortex V4.","PeriodicalId":314049,"journal":{"name":"Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115239753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Feature selection is one of the most important data mining research topics with many applications. In practical problems, features often have group structure to effect the outcomes. Thus, it is crucial to automatically identify homogenous groups of features for high-dimensional data analysis. Octagonal shrinkage and clustering algorithm for regression (OSCAR) is an important sparse regression approach with automatic feature grouping and selection by ℓ1 norm and pairwise ℓ∞ norm. However, due to over-complex representation of the penalty (especially the pairwise ℓ∞ norm), so far OSCAR has no solution path algorithm which is mostly useful for tuning the model. To address this challenge, in this paper, we propose a groups-keeping solution path algorithm to solve the OSCAR model (OscarGKPath). Given a set of homogenous groups of features and an accuracy bound ε, OscarGKPath can fit the solutions in an interval of regularization parameters while keeping the feature groups. The entire solution path can be obtained by combining multiple such intervals. We prove that all solutions in the solution path produced by OscarGKPath can strictly satisfy the given accuracy bound ε. The experimental results on benchmark datasets not only confirm the effectiveness of our OscarGKPath algorithm, but also show the superiority of our OscarGKPath in cross validation compared with the existing batch algorithm.
{"title":"Groups-Keeping Solution Path Algorithm for Sparse Regression with Automatic Feature Grouping","authors":"Bin Gu, Guodong Liu, Heng Huang","doi":"10.1145/3097983.3098010","DOIUrl":"https://doi.org/10.1145/3097983.3098010","url":null,"abstract":"Feature selection is one of the most important data mining research topics with many applications. In practical problems, features often have group structure to effect the outcomes. Thus, it is crucial to automatically identify homogenous groups of features for high-dimensional data analysis. Octagonal shrinkage and clustering algorithm for regression (OSCAR) is an important sparse regression approach with automatic feature grouping and selection by ℓ1 norm and pairwise ℓ∞ norm. However, due to over-complex representation of the penalty (especially the pairwise ℓ∞ norm), so far OSCAR has no solution path algorithm which is mostly useful for tuning the model. To address this challenge, in this paper, we propose a groups-keeping solution path algorithm to solve the OSCAR model (OscarGKPath). Given a set of homogenous groups of features and an accuracy bound ε, OscarGKPath can fit the solutions in an interval of regularization parameters while keeping the feature groups. The entire solution path can be obtained by combining multiple such intervals. We prove that all solutions in the solution path produced by OscarGKPath can strictly satisfy the given accuracy bound ε. The experimental results on benchmark datasets not only confirm the effectiveness of our OscarGKPath algorithm, but also show the superiority of our OscarGKPath in cross validation compared with the existing batch algorithm.","PeriodicalId":314049,"journal":{"name":"Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115849459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：